1. Field of the Invention
This invention relates to a contents data recording/reproduction apparatus that generates index data for the contents data, and generates contents data based on the generated index data.
2. Description of the Related Art
In recent years, as recording apparatuses such as a VTR (video tape recorder) that records or reproduces data such as that of a program from a television broadcast have become widely spread, performing a summary reproduction that provides a summary in a short period of time of the contents data without having to watch all of the recorded contents has been employed.
Conventionally, as methods for performing this kind of summary reproduction, there is a method of performing summary reproduction by paying attention to the audio data that accompanies the contents data, and a method of performing summary reproduction by mainly paying attention to the video data itself. Particularly, the summary reproduction method that performs summary reproduction by paying attention to the video data itself is such that it recognizes telop text that indicates the contents of the time progression that are presented in the main image data of the video, and identifies the contents of the images of each part of the main image data based on that recognized text, then performs a summary reproduction based on the identified partial image data (see; Institute of Image Electronics Engineers of Japan Paper 02-07-33, “Auto-Indexing and Summary Generation of the Baseball Season Driven by Telop”)
However, in the conventional method of performing summary reproduction using telop, a plurality of partial image data in the main image data is identified based on information in a recognized telop, and an index for the contents data is set based on the identified partial image data.
Therefore, in that summary method, the relationships between partial image data, such as the importance occupied in the entire video data by each partial image data that was identified by the index, were not displayed, and it was not possible to easily search for arbitrary scenes in the contents data desired by the user based on the generated index.